RESTful Software Architecture for ROS-based Onboard Mission System for Drones

Abstract

Unmanned Aircraft Systems (UAS), commonly known as drones, have recently gained much attention and are being deployed in many fields such as delivery services, surveillance, and agriculture. Many drone applications require autonomy features and ask for integration into other platforms. This paper investigates software architecture for drones that fosters interoperability with web services while supporting autonomy. It extends an existing autonomy architectures with a web service layer that allows task level integration through a Representational State Transfer (REST) interface. Underlying Robot Operating System (ROS) layer not only enriches the architecture with the capabilities of the available ROS nodes, but also provides a state of the art middleware support. The proposed architecture is demonstrated in Gazebo where a simulated drone is flown with an onboard mission system on a Raspberry Pi connected remotely to the RESTful client, thereby allowing remote execution of missions through web services.